m-clark.github.io/docs/sem/sem_workshop.zip
Steps:
For exercises, open the exercises_nb.nb.html
with your browser!
Or go to:
Preliminaries
Graphical Modeling
Latent Variables
SEM
Overview of others
Give a broad overview of techniques related to SEM
Note connection to others
Be clear on limitations
Primary components
Scripts/Docs
Console
Viewer
Other
For this workshop, use traditional R script or notebook
Script is where your code goes
Ctrl+Enter will run current line or selected lines
Where the code is run
Typically your output will be viewed here
Avoid using directly
Visualizations
Other HTML
RStudio tries to make things easier
Let it!
Objects, Functions
Classes
Installing packages
Using functions to create output
Print the object to view it
myobj = function(arg1=input1, argA=inputA)
myobjThe following creates the object nums
sum sums the elements of nums
The result is assigned to the object numsSum
Object is printed to view it
nums = c(1,2,3,4) # c is a function that combines its elements
class(nums)[1] "numeric"
numsSum = sum(nums)
numsSum[1] 10
library(lavaan) # loads the library
myModel = "
# latent variables aka measurment models
Factor1 =~ x1 + x2 + x3
Factor2 =~ y1 + y2 + y3
# regresssions aka structural model
Factor2 ~ Factor1 + z1 + z3
"
mySEM = sem(myModel, otherinputs) # run the model
summary(mySem) # display resultsAlways save your scripts
If desired, can save your work as RData file
save(obj1, obj2, file='filelocation/myRstuff.RData')
save.image('filelocation/myRstuff.RData') # saves everythingDirected Graphs
Path Analysis
Mediation
Bayesian Networks
Undirected
Network analysis
Nodes, vertices
May define variables, observations, concepts
Edges, links
Define connections, relations
Directed, Undirected, Mixed
Extends regression to:
Multiple targets
Indirect effects
Does not require SEM
Common in some areas:
Indirect effects
Still may include direct path
Always more complicated than three variables
Very difficult to justify in cross-sectional setting
Strong causal implications
Does not prove a causal relationship
Bayesian Networks
Network Analysis
Data Compression
Measurement of Latent Constructs
Other
Take many things and produce few
Might focus on variance (e.g. PCA) or covariance (FA)
Use estimated scores in other analysis
Underlying causal model
\[ x = \lambda \textrm{F} + \epsilon \]
\[ \epsilon = \textrm{systematic + measurement error} \]
Psychological constructs
Scholastic abilities
Topics in text
Number of factors
Loadings, interpretation
Scale development, reliability
Many, including:
How to assess fit
Identification
Estimation techniques
Model complexity
Sample Size
Too little data
‘Poor’ data
Ignoring performance, diagnostics
Latent Classes
Growth Curves
Multi-sample
Because it sounds cool
Because you see it in a paper
Because you can’t get a large sample
It’s can answer your specific research question better than alternatives
Very strong result if it works
Measurement is of focus
Ties to causal thinking
Flexibility with Complexity